A pilot study to compare natural health product–drug interactions in two databases in Canada

Objective The aim of this article is to evaluate and compare two natural health product databases for the purpose of integrating them into a pharmacy information system in Canada. Methods This is a descriptive pilot study that compared the Natural Medicine Comprehensive Database® (NMCD) and Natural Standard® Database (NSD). We randomly chose five hospital patient files. For each drug prescription, we searched for and counted the number of potential natural health product–drug interactions in each database. Main outcome measure We compared all of the potential interactions between dexamethasone and the natural health products in both databases. We also evaluated the quality of a selection of references for 30 potential natural health product–dexamethasone interactions. Results Five pediatric patient files were selected for a total of 21 different common names. The number of potential natural health product–drug interactions identified varied from 12 (salbutamol-albuterol) to 129 (dexamethasone) in the NSD for an average of 63 ± 33. The number of potential natural health product–drug interactions identified varied from 1 (salbutamol-albuterol) to 96 (dexamethasone) in the NMCD for an average of 55 ± 27. There was no significant difference between the average number of potential natural health product–drug interactions between the databases (P = 0.40). The average number of common potential interactions was 9 ± 8. Thirty potential interactions of dexamethasone were compared. The number of abstracts per pair of interactions varied from 0 (Aloe-dexamethasone in the NSD) to 17 (dexamethasone-St. John’s wort in the NMCD). For the 10 pairs that were common to both databases, the number of abstracts given was 4.5 ± 4.7 for the NMCD and 3.1 ± 2.1 for the NSD (P = 0.41) and the average evaluation score was 4.6 ± 1.5 for the NMCD and 5.6 ± 1.2 for the NSD (P = 0.43). Conclusion This pilot study presents a comparison of potential natural health product–drug interactions from two different databases using two different scenarios in a hospital setting. The study reveals a high and variable number of pairs of natural health product–drug interactions per drug between the NSD and the NMCD. Further studies are necessary to evaluate the pertinence and hierarchy of the information offered by suppliers and natural health product databases.